not accurate results from Moco Track

[Nicholas Bianco]

Hi Mohammad,

Are the problems you run in 4.2/4.3 actually converging? The "constant" ankle angle result might be that the problem is failing on the first iteration, where the initial guess is the midpoint value between the bounds.

Perhaps you could provide the log files for the 4.2/4.3 solution so I can see error messages and anything else that may be useful for debugging here.

-Nick

[Mohammad RahimiGoloujeh]

Hi Nick,
i was not sure how to get the log file. i just ran the code from conda prompt and copy and paste the solver log info
hope that helps
i really appreciate your help.

[Nicholas Bianco]

Hi Mohammad,

Well it looks like your problem is converging. I see you're applying external loads to the model, are these used for your platform perturbation? Or are you using a prescribed motion for that? It would be helpful to explain your model and the problem you are trying to solve so I can help troubleshoot.

-Nick

[Mohammad RahimiGoloujeh]

Hi,
The external forces are ground reaction forces from Vicon.
for the perturbation I prescribed the platform motion based on experimental data ( the prescribed motion is compatible with the data i want to track)

[Nicholas Bianco]

Hi Mohammad,

Is it possible to track joint coordinates and velocities rather than marker data? Another user on the forum has been having issues with marker tracking problems (including a Moco example), something might be broken in the MocoMarkerTrackingGoal.

Best,
-Nick

[Mohammad RahimiGoloujeh]

Hello,
Thank you for your reply,
I am going to get the inverse solution based on the markers and use the resultant joint coordinates to track and will tell you the results
thank you

[Nicholas Bianco]

Hi Mohammad,

I recently found a bug related to marker tracking MocoTrack. If this was causing your problems, you can find a workaround for it on this post: viewtopicPhpbb.php?f=1815&t=14612&p=42525&start=0&view=.

Best,
Nick